Search

Introduction Sleep disruption, and in particular decreased NREM stage N3 sleep, has been implicated in cognition and pathogenesis of Alzheimer's Disease (AD). Several recent findings suggest that enhancing N3 sleep may mitigate impairments due to chronic sleep disruption. However, approaches such as transcranial electrical stimulation (TES) have been limited by discomfort, awakenings, and lack of blinding. In response to the need to develop low-level TES, our team first identified the source of slow oscillations (SOs) of N3 sleep using high-density EEG array modeling (Sourcerer®), which localized within anterior limbic brain structures. Next, we showed that low-level TES targeted to this region induced and enhanced SOs during N3 sleep in the laboratory. More recently, we developed a portable TES device with the minimum deployable set of stimulation electrodes targeting the anterior limbic SO-generating regions: Wireless Interface Sensor Pod (WISP). Methods Thirteen healthy adults (m = 42 years) participated in three all-night sleep EEG recordings where they received low level (0.5 mA) TES designed to target anterior limbic SO-generating regions and a sham stimulation (placebo). TES was initiated after 4 minutes of naturally occurring NREM N2 sleep and included five TES blocks of five minutes each separated by one minute rest periods. Subjects were blinded to stimulation and sham conditions, and outcomes included feasibility, acceptability, and total time in each sleep stage. A convolutional neural network (CNN) was developed, trained, and tested on manually scored EEG sleep staging for automated sleep scoring. Results Subjects rated that the WISP device was highly feasible and acceptable. Subjects were adequately blinded to sham versus stimulation. When compared to the sham session, limbic-targeted TES significantly increased the duration of N3 sleep and spectral power in the 0.5-1 Hz frequency band. Finally, our CNN was able to detect sleep stages with good accuracy (84%, kappa 0.77). Conclusion These results suggest that low-level portable TES using a minimum deployable set of electrodes, when specifically targeting anterior limbic sites, is feasible, acceptable, maintains blinding, and increases deep (N3) sleep. The WISP device is a promising method to both assess and manipulate sleep in the home setting. Support (if any) DoD MOMRP MTEC-21-06-MPAI-129

Introduction Prospective studies of Alzheimer’s disease (AD) demonstrate sleep-wake disturbances may precede and accelerate with cognitive decline. Use of wrist-actigraphy in conjunction with overnight polysomnography (PSG) enables researchers to evaluate diverse characteristics of sleep and circadian rhythms. Though outcome measures from these devices have been implicated in the progression of AD, because of great individual differences, it is valuable to internally replicate within cohort. Here we demonstrate application of actigraphy in the context of healthy aging to identify patterns of activity that predict cognitive decline. Methods Participants were prospectively recruited at Washington University as part of the BASE study and wore an Actiwatch2 (Philips-Respironics) and cross-validated for sleep and wake using polysomnography. The watch was worn continuously for at least 5 consecutive nights to be included in the analysis (n = 68, age = 71.1 ± 4.5 years). Rest-activity rhythms were computed using compiled code from National Sleep Research Resource (actiCircadian) pipeline implemented in MATLAB R2021a (Mathworks). Clinical Dementia Rating (CDR) scales and cognitive tests were conducted on the first day of actigraphy recording. Results Individuals with cognitive impairment (CDR Global Score > 0) had increased wake after sleep onset (WASO) and reduced sleep efficiency compared with individuals with no impairment (CDR = 0). (unpaired ttest: p = 0.02 and p = 0.02). Participants with cognitive impairment had a trend toward reduced relative amplitude (RA) (p = 0.06). Reduced (worse) RA was correlated with worse performance on the Trail Making Task A (Pearson r: -0.37, p = 0.002). Higher (better) interdaily Stability (IS) was associated with better responses on subjective questionnaires including Epworth Sleepiness Scale and Pittsburgh Sleep Quality Index (Pearson r: -0.24 p = 0.051; r: 0.242, p = 0.047, respectively). Conclusion Actigraphically-derived rest-activity metrics are correlated with cognitive status, cognitive test performance, and subjective sleep questionnaires. Support (If Any) NIH R01 AG059507, ARCS Foundation Scholar

Introduction There exists a bidirectional relationship between sleep disruption and neuropathology in Alzheimer’s disease (AD). Electroencephalogram (EEG) during polysomnography (PSG) provides an opportunity to examine stereotyped, coordinated brain activity. Specifically, slow wave activity (SWA), a defining feature of non-REM sleep, is aberrant in AD, and disruption of SWA in healthy adults is related to increased amyloid-beta levels. Coherence of slow waves across different cortical regions may serve as a metric of brain network coordination, sensitive to earliest AD pathology. We explored slow wave coherence during sleep in relation to sleep macrostructure and cognitive performance in the Biomarkers of Alzheimer’s Disease in Sleep and EEG(BASE) cohort. Methods EEG was collected during an attended overnight PSG from the BASE cohort (n=79, average age = 70.8 ± 4.4 years), approximately 20% of whom had Clinical Dementia Rating (CDR) of 0.5-1. A custom slow wave peak detector was implemented in MATLAB to across 6 EEG leads (C3, C4, F3, F4, O1, and O2), and slow wavecoherence was calculated as the proportion of sleep with slow waves occurring within 100ms in all 6 leads. Results EEG slow wave coherence was reliably quantified during wake, non-REM stages N1, N2, N3, and REM. Slow wave coherence in N2/N3 was negatively correlated with measures of sleep fragmentation (wake after sleep onset Spearman r = -0.3, p 0.008; number of awakenings r = -0.39, p 0.0004), and positively correlated with sleep efficiency (r = 0.28, p = 0.013). Slow wave coherence in N2/N3 was correlated with overall cognition as measured by MoCA adjusted z-score (r = 0.2, p 0.087) and executive function as measured by Trailmaking B adjusted z-score (r = 0.25, p 0.038). Conclusion Slow wave coherence was strongly correlated with measures of sleep quality. Reduced slow wave coherence was associated with poorer cognitive executive function. EEG slow wave coherence is a novel, promising approach to measure coordinated brain network function that is sensitive to early cognitive dysfunction in AD. Support (If Any) NIH R01 AG059507, ARCS Foundation Scholar

Introduction Chronic pain and sleep disturbances are intricately linked to one another, especially in individuals with a history of traumatic brain injury (TBI) who are at greater risk for both symptoms. Although prior studies have analyzed differences in sleep electroencephalogram (EEG) in these clinical populations, the association between sleep EEG slow wave coherence and pain complaints is not fully examined or known. Our novel slow wave coherence approach may provide new insights into the relationship between TBI, chronic pain, and sleep Methods Ninety-six veterans were recruited and enrolled under a VA IRB-approved protocol. Participants completed a semi-structured clinical interview to determine their history of TBI, Symptom Impact Questionnaire Revised (SIQR), a measure of chronic pain complaints, and underwent an attended overnight in-lab polysomnogram (PSG). We developed a novel computational signal processing algorithm to identify and quantify EEG slow waves within 100 ms bins across the 6 standard PSG EEG channels. When a slow wave was simultaneously observed in 4 or more of the 6 leads, slow wave coherence was inferred, and a percentage of slow wave coherence across each of the sleep stages was then calculated for each subject. Results In our sample, 65 participants (67.7%) endorsed experiencing chronic pain lasting 3 months or longer, and 54 had a history of TBI (56.3%). Participants endorsing chronic pain had a significantly lowered percent of EEG slow wave coherence during NREM sleep than subjects without chronic pain (p = 0.01). NREM EEG slow wave coherence did not correlate with SIQR scores in subjects without TBI (r = -0.03, p = 0.90), but was significantly negatively correlated in subjects with TBI (r = -0.32, p = 0.02). Conclusion EEG slow wave coherence during NREM sleep is correlated with chronic pain complaints in Veterans with a history of TBI, and could be indicative of neuronal dysfunction during sleep. Further research on slow wave coherence is warranted to understand the underlying mechanisms for the association between chronic pain and poor sleep following TBI. Support (if any) D01 W81XWH-17-1-0423

Introduction Glutamate concentrations in the cortex fluctuate with the sleep wake cycle in both rodents and humans. Altered glutamatergic signaling, as well as the early life onset of sleep disturbances have been implicated in neurodevelopmental disorders such as autism spectrum disorder. In order to study how sleep modulates glutamate activity in brain regions relevant to social behavior and development, we disrupted sleep in the socially monogamous prairie vole (Microtus ochrogaster) rodent species and quantified markers of glutamate neurotransmission within the prefrontal cortex, an area of the brain responsible for advanced cognition and complex social behaviors. Methods Male and female prairie voles were sleep disrupted using an orbital shaker to deliver automated gentle cage agitation at continuous intervals. Sleep was measured using EEG/EMG signals and paired with real time glutamate concentrations in the prefrontal cortex using an amperometric glutamate biosensor. This same method of sleep disruption was applied early in development (postnatal days 14–21) and the long term effects on brain development were quantified by examining glutamatergic synapses in adulthood. Results Consistent with previous research in rats, glutamate concentration in the prefrontal cortex increased during periods of wake in the prairie vole. Sleep disruption using the orbital shaker method resulted in brief cortical arousals and reduced time in REM sleep. When applied during development, early life sleep disruption resulted in long-term changes in both pre- and post-synaptic components of glutamatergic synapses in the prairie vole prefrontal cortex including increased density of immature spines. Conclusion In the prairie vole rodent model, sleep disruption on an orbital shaker produces a sleep, behavioral, and neurological phenotype that mirrors aspects of autism spectrum disorder including altered features of excitatory neurotransmission within the prefrontal cortex. Studies using this method of sleep disruption combined with real time biosensors for excitatory neurotransmitters will enhance our understanding of modifiable risk factors, such as sleep, that contribute to the altered development of glutamatergic synapses in the brain and their relationship to social behavior. Support (if any) NSF #1926818, VA CDA #IK2 BX002712, Portland VA Research Foundation, NIH NHLBI 5T32HL083808-10, VA Merit Review #I01BX001643

Introduction Sleep disturbances are common in elderly patients and may contribute to disease progression in certain populations (e.g., Alzheimer’s Disease). Light therapy is a simple and cost-effective intervention to improve sleep. Primary barriers to light therapy are 1) poor acceptability to use of devices and 2) inflexibility of current devices to deliver beyond a fixed spectrum and throughout the entirety of the day. However, dynamic, tunable lighting integrated into the native home lighting system can maximize short-wavelength light in the morning and minimize short-wavelength light in the evening, thus entraining circadian rhythms and treating sleep disturbances, and overcome these limitations. We determined the feasibility of implementing a whole-home tunable lighting system as a potential sleep intervention. Methods Tunable LED lights were installed throughout the homes of healthy older adults already enrolled in an existing study with embedded home assessment platforms (ORCATECH study; n=4 subjects in n=3 homes). In ORCATECH, continuous data on room location, activity, sleep, and general health parameters are collected at minute-to-minute resolution over months to years of participation. This single arm longitudinal design collected participants’ light usage in addition to ORCATECH outcome measures. Primary outcomes for this pilot study included the feasibility and patient acceptability. Exploratory outcomes were sleep metrics (sleep time, latency, efficiency), mobility (room transitions and actigraphy), and overall health indices (weekly body weight, self-report general health questionnaires) both pre- and post-intervention. Results Two subjects terminated the study citing technical difficulties with the lights and a preference for brighter illumination. Of the remaining 2 participants, sleep metrics were explored over a 12-month period spanning pre- and post-installation of lights. Nightly duration in bed was compared with minute-to-minute room entry data and actigraphy with high inter-measure reliability. Conclusion These data support that tunable whole-home lighting systems are reasonably acceptable and feasibly implemented using an automated platform for continuous data collection. Quantification of sleep over long periods of time is robust and reliable in the home environment of elderly subjects. These results will inform implementation of future large-scale lighting intervention studies in patients at risk for developing Alzheimer’s Disease. Support (if any) Hartford Gerontological Center Interprofessional Award, Pacific Northwest National Laboratory, OHSU ORCATECH

Introduction A bidirectional relationship exists between sleep disruption and neuropathology in Alzheimer’s disease (AD). The sleep electroencephalogram (EEG) is a highly stereotyped, direct neurophysiological window into brain function; prior studies have identified abnormalities in EEG slow waves in early AD. EEG coherence across channels during sleep, a normally highly coherent brain state, could be an indicator of network coordination across brain regions. Accordingly, altered slow wave coherence during sleep may be an early indicator of cognitive decline. Methods EEG was collected during an attended overnight polysomnogram (PSG) from a community-based cohort of older subjects (n=44, average age = 71), approximately 25% of whom met criteria for mild cognitive impairment or early AD. Files were exported to EDF and a slow wave peak detector was implemented in MATLAB to count the number of slow wave oscillations, with automated artifact rejection, across 6 EEG leads standard for PSG (C3, C4, F3, F4, O1, and O2). Slow wave coherence was inferred when slow waves occurred in temporal synchrony across channels within 100 ms. Results Subjects with cognitive impairment showed significantly reduced total sleep time and time spent in rapid eye movement (REM) sleep compared to age-matched controls. EEG slow wave coherence was reliably quantified during wake, non-REM stages N1, N2, N3, and REM vigilance states as well as during transition periods between sleep stages. Using this algorithm, specific signatures of slow wave propagation during sleep were identified, including increased variability in slow wave activity and coherence, that appeared more prominent in subjects with impaired cognition. Conclusion EEG slow wave coherence during sleep and wake states can be calculated by applying automated algorithms to PSG data, and may be associated with cognitive impairment. Support (if any) NIH R01 AG059507